Alan0354
Full Member level 4
Hi
I have a 3 stage gain circuit each with gain of 10. I use Ti THS2301 that has about 320MHz BW at gain of 10. The circuit split into two pcb connected together by short coax.
1) First pcb has two stages driving the coax to the second pcb. input is a 50ohm coax
2) Second pcb has one stage of gain=10 before driving the LT6752 comparator with CMOS output.
The symptom is the input coax is very sensitive to position, I suspect is the CMOS output feeding back to the coax. It is not necessary to the first stage because I put shield over the first stage amplifier and it does not help too much.
I finally put a toroid on the input coax close to the termination end. That helps a lot. I then put another toroid at the CMOS output coax, that together with the toroid on the input coax, that fixed the oscillation, I can adjust the threshold much lower on the comparator.
I think it's the output is so high ( 2V with 1nS rise time) that couple to the input coax. I want to know any other way to improve this and possibility getting rid of the toroid.
You think if I go to comparator with differential output, that would help as the total current cancel and much less radiation out of the output line.
thanks
I have a 3 stage gain circuit each with gain of 10. I use Ti THS2301 that has about 320MHz BW at gain of 10. The circuit split into two pcb connected together by short coax.
1) First pcb has two stages driving the coax to the second pcb. input is a 50ohm coax
2) Second pcb has one stage of gain=10 before driving the LT6752 comparator with CMOS output.
The symptom is the input coax is very sensitive to position, I suspect is the CMOS output feeding back to the coax. It is not necessary to the first stage because I put shield over the first stage amplifier and it does not help too much.
I finally put a toroid on the input coax close to the termination end. That helps a lot. I then put another toroid at the CMOS output coax, that together with the toroid on the input coax, that fixed the oscillation, I can adjust the threshold much lower on the comparator.
I think it's the output is so high ( 2V with 1nS rise time) that couple to the input coax. I want to know any other way to improve this and possibility getting rid of the toroid.
You think if I go to comparator with differential output, that would help as the total current cancel and much less radiation out of the output line.
thanks
